Pumped hydro — storing by lifting water

Reservoirs, pumps, turbines, land, losses and a long service life. Theory, formulas, worked examples and applications.

Storing electricity by lifting water

The principle: turning electricity into height, and height into electricity

Pumped hydro storage is, conceptually, the oldest and simplest form of large-scale electricity storage: when the grid has surplus electricity — early in the morning when demand is low, or at midday when solar generation is abundant, for example —, electric pumps lift water from a lower reservoir to an upper one, converting electrical energy into gravitational potential energy. When that energy is needed again, the water flows down through turbines back to the lower reservoir, returning part of the stored energy as electricity. The system never creates electricity out of nothing at any point: less energy is always recovered than was invested in pumping, because every conversion — electrical to mechanical, mechanical to potential, and back — has its own losses from friction, turbulence and machine efficiency. What the system really sells is not additional net energy, but the ability to move energy across time: storing it when there is a surplus, delivering it when there is a shortfall.

What has to be built to lift and release water at scale

Building a pumped-hydro facility requires artificial reservoirs or tanks — often two, an upper and a lower one, separated by a significant geographic elevation difference —, large-scale excavation, in many cases tunnels connecting both reservoirs through a mountain, large amounts of concrete and steel for dams and structures, penstocks able to withstand enormous pressures, reversible turbines that work both as pump and as generator, grid connection lines, and access roads for construction and maintenance. A closed system, reusing the same water between the two reservoirs almost indefinitely, only needs to replace evaporation and small leakage losses. A system connected to a river, by contrast, interacts much more directly with natural flow, sediments and river ecosystems, similarly to any other conventional hydroelectric infrastructure.

Topography decides where it is possible, not the will to build

Unlike a chemical battery, which can in principle be installed on any flat site, a pumped-hydro facility is only viable where geography offers a suitable elevation difference between two reasonably close points — two valleys, a mountain with a usable slope, or an already-excavated quarry or mine that can be converted. That geographic dependence drastically limits the number of genuinely available sites in any country or region, and explains why pumped hydro, despite being the world's most widespread grid-storage technology, cannot simply be deployed wherever demand requires it, only where the terrain allows it.

Advantages of scale and lifespan, disadvantages of land and time

Pumped hydro's great advantage is its ability to operate at a much larger scale than any current chemical battery — hundreds or thousands of megawatts, able to deliver that power for many hours in a row — and with an exceptionally long service life: dams and tunnels can keep operating for many decades with proper maintenance. Its disadvantage is equally clear: it requires considerable land transformation, years of construction time, and is limited to a small number of geographically suitable sites. Pumps and turbines are periodically renewed over the facility's life, but dams, tunnels and slopes require continuous structural monitoring for decades, because a structural failure at a dam has potentially catastrophic consequences, as developed in the hydropower leaf in this platform's Energy Sources section. At the end of its useful life, what remains is not a simple empty container that can just be removed: it is a large hydraulic infrastructure permanently integrated into the landscape.

Comparing with batteries: the right question is not "which is better?"

It makes no sense to ask in the abstract whether pumped hydro or chemical batteries are "the best" storage technology: the answer depends entirely on what specific service is needed. One has to compare the instantaneous power each option can deliver, how many hours of continuous autonomy it offers, how many charge-discharge cycles it withstands before degrading, the construction time required, the land it occupies, the water and materials it requires, its response speed to sudden demand, and its total expected lifespan. Chemical batteries are modular, install in weeks or months and respond in milliseconds, making them ideal for fast, short-duration services. Pumped hydro can last far longer and store much larger amounts of energy where geography allows, but requires years of planning and construction before delivering its first megawatt-hour.

Geography decides before engineering does

In favor: the most mature, largest-scale grid storage technology that exists, with a service life of many decades and the ability to deliver power for hours on end — something no chemical battery yet matches at that volume.

To watch: only viable where topography offers a suitable elevation difference between two nearby points; requires years of construction and considerable land transformation, with dams and slopes requiring structural monitoring throughout their whole service life.

Do not ask which is better in the abstract. Compare instantaneous power, hours of autonomy, cycles withstood, construction time, land occupied, water and materials needed, response speed and expected lifespan. Batteries are modular and quick to install; pumped hydro can last much longer and store much larger amounts where geography allows.

Before proposing a new facility: always compare with converting an existing quarry, an exhausted mine, or an already-existing hydroelectric installation — it usually requires far less new land transformation than building two reservoirs from scratch in a still-intact valley.

The numbers, with their source

More than 90% of all the world's grid-scale storage

Pumped hydro storage represents, by various estimates, more than 90% of all grid-scale electricity storage capacity installed worldwide, and more than 94% of long-duration storage. Global installed capacity surpassed 200 GW in 2025, with another 243 GW under construction across different countries — a volume that, despite the recent boom in large-scale batteries, remains by far the planet's dominant electricity storage technology.

Sources: Hydropower.org — Year of the water battery: global pumped storage capacity surpasses 200 GW; Hydropower.org — Pumped storage hydropower explained.

Why it dominates despite all its geographic limitations

That a technology limited to a small number of geographically suitable sites still represents the vast majority of the world's storage says a lot about the real scale required to store electricity for an entire grid: chemical batteries, though growing rapidly, still represent a small fraction of that total capacity, precisely because building batteries at the gigawatt-hour scale offered by a large pumped-storage reservoir remains, today, far more expensive per unit of energy stored for many hours.

What this table is missing

This installed-capacity figure does not distinguish between facilities that have operated for many decades — built when the opportunity cost of land was different from today's — and new projects that now have to compete with chemical batteries, whose cost has fallen rapidly in recent years; comparing an existing facility's cost with a new one's without that distinction is misleading.

Sources